Historical temperature reconstructions show hemispheric differences

Figure detailing the warm and cold periods reconstructed for the northern and southern hemispheres
These figures show the warm and cold periods reconstructed from proxy records (red and blue) in the Northern Hemisphere (a) and Southern Hemisphere (b). Model simulations (pale colours) are also superimposed over the reconstructions. The Medieval Warm Period (950-1250AD) is apparent in the Northern Hemisphere reconstruction, while in the Southern Hemisphere there was an extended warm phase between 1200 and 1350. Both hemispheres then experienced a long-term cooling trend, with the peak of the Little Ice Age from 1594-1677. Since the 1970s both hemispheres have experienced synchronised warming extremes. Figure c shows simultaneous extreme periods in both hemispheres. Some of the model simulation results deviate from the reconstructions at times when external forcings (volcanic eruptions, solar radiation, greenhouse gases) were more extreme (d), as a result of the models’ over-emphasis on these forcings. The axis ‘fraction of ensemble members’ refers to the number of proxy records that show the warming or cooling trend. (First published in Nature Climate Change doi: 10.1038/NCLIMATE2174).

Climate scientists have found that temperature variations in the northern and southern hemispheres over the past 1000 years are not as synchronised as previously thought, with implications for future climate modelling.

The research, published in Nature Climate Change in April, found that Southern Hemisphere temperature fluctuations over the past millennium showed marked differences to that of the more intensively studied Northern Hemisphere.

Until now, it was thought that Southern Hemisphere climate mirrored that of the north as a result of external factors that affect climate, such as solar radiation, greenhouse gases and volcanic eruptions.

However, Australian Antarctic Division climate program leader, Dr Tas van Ommen, who was involved in the internationally collaborative research, said ‘internal’ ocean–atmosphere dynamics, which differ in both hemispheres, had more of an influence on Southern Hemisphere climate than previously thought.

‘Our analyses suggest that models may be underestimating the role of internal ocean-atmosphere dynamics in driving short-term climate variability, particularly in the ocean-dominated south. So while model projections may capture the long-term influence of external climate forces, they would be missing the full impact of short-term variations – up to a few decades long – above or below this trend,’ Dr van Ommen said.

‘This means that some of the warming or cooling phases previously attributed to both hemispheres, did not occur in the south.’

The team, led by Swiss researcher Raphael Neukom, reconstructed yearly temperature variations using an extensive database of terrestrial and oceanic palaeoclimate ‘proxy’ records from the Southern Hemisphere. Proxies are physical records that were influenced by the climate in which they formed – in this case the team used tree rings, marine and lake sediments, ice cores, corals and cave structures, as well as documented climate records from post-1970s.

‘Our proxy records came from more than 300 individual sites, nearly doubling the number of records considered in the most advanced previous reconstruction attempt,’ said study co-author Dr Tessa Vance, of the Antarctic Climate & Ecosystems Cooperative Research Centre.

‘This allowed us to develop a well verified Southern Hemisphere temperature reconstruction for every year of the past millennium.’

The team compared this to an independent Northern Hemisphere temperature reconstruction. They found that both the northern and southern hemispheres shared an extended cold period between 1571 and 1722 (known as the ‘Little Ice Age’), but there were differences in the existence, timing and phase of warm and cold periods throughout the rest of the millennium.

The most striking difference was the absence in the Southern Hemisphere of the Medieval Warm Period, between 950 and 1250 AD.

‘This means that the Medieval Warm Period, which was previously observed in Northern Hemisphere-centric reconstructions of temperature, is not, as is often suggested, a worldwide phenomenon, ‘ Dr Vance said.

‘In fact, there is no globally synchronised warm phase during the pre-industrial era, between 1000 and 1850.

However, since the 1850s both hemispheres have experienced synchronous warming, with modern warming extremes evident from 1979 in more than 90% of the proxy records used.

The new temperature reconstructions for the Southern Hemisphere suggest that data from the Northern Hemisphere alone are insufficient to characterise global scale temperature anomalies, trends and extremes.

These findings need to be factored into future climate simulations to ensure the most robust predictions for regional and hemispheric climate.

‘Analyses targeting periods where climate model simulations and reconstructions differ, will be necessary to identify weaknesses in both proxy- and model-based representations of the Earth’s climate system,’ Dr van Ommen said.

‘Other studies of Southern Hemisphere climate are also needed to gain a better understanding of inter-hemispheric differences.’

Neukom R, Gergis J, Karoly D, et al (2014). Inter-hemispheric temperature variability over the past millennium. Nature Climate Change doi: 10.1038/NCLIMATE2174.

Wendy Pyper
Corporate Communications, Australian Antarctic Division